Opinion
AI

Turning AI Use into an Operating and Strategic Framework for Destinations

AI only becomes an organisational capability once it is applied consistently. Aruba Tourism Authority's AI strategy was built around that belief, replacing scattered, individually driven AI use with a single governed approach.

In this piece

AI adoption rarely stalls for lack of enthusiasm. Staff find their own way to tools that make daily work easier and momentum builds long before any formal guidance catches up. The more significant test comes afterwards, when an organisation has to decide what to do with that momentum. Left to develop on its own, individual experimentation produces expertise that stays with the person who gained it, tools that were never risk assessed and a picture of AI use that leadership teams are unable to observe.

This was the position the Aruba Tourism Authority (A.T.A.) found itself in when it brought in the Digital Tourism Think Tank (DTTT) for a strategic AI engagement. A.T.A.'s work spans destination management, trade development, visitor experience and research, carried out by a workforce that stretches from its Aruba headquarters to market teams around the world.

AI was already part of internal workflows. Some departments had already built their own AI tools, while an AI-powered visitor-facing chatbot called myAruba Assistant was live and had already handled more than 10,000 conversations. Everywhere else, AI use was individually driven and no one had a shared view of what was being used or how. DTTT was commissioned to turn that energy into a single, governed approach, consisting of an audit of where things stood, a strategy for how AI should be leveraged and a roadmap for getting there.

Enthusiasm Without a Shared Foundation

DTTT's approach began with consultations across headquarters and market teams, combining strategy sessions, hands-on workshops and one-to-one departmental conversations. A theme only made it into the strategy once it had shown up independently across several of these settings, so what reached the strategy reflected experience shared consistently across the organisation. That evidence then shaped three deliverables that built on one another in sequence. The audit set out what the research had found, the strategy translated those findings into a clear direction and the roadmap turned that direction into action, with each recommendation carried through into a specific, owned task.

Research carried out across the entire A.T.A. team found a workforce that had taken to AI faster than the organisation had built the structure to support it. 91% of staff were already using AI tools in their daily work, yet only 8% felt confident doing so. Closing that gap between how widely AI was being used and how confident people felt using it became the starting point for everything the strategy set out to do.

The audit pointed to a clear opportunity here. Organising the way information is stored and connected across the organisation would let staff draw on AI with far greater confidence, trusting that the answers a tool gives back are grounded in the right source material. Put simply, well-organised knowledge is what turns AI from something staff use cautiously into something they can rely on.

Staff were also thinking carefully about how data was being handled. What came through clearly was a strong desire for structure, with staff wanting access to the right tools, provided by the organisation itself instead of relying on shadow AI use, backed by clear guidelines and content standards that give them the confidence to use AI well.

AI maturity also varied across the organisation, with some departments further ahead than others in integrating AI into their work. Any AI literacy programme had to start from each team's actual position rather than assume a single starting line for everyone. What made the picture encouraging was the appetite sitting underneath it, with 80% of staff willing to commit one to two hours a week to developing their AI skills further. Development pathways, tool recommendations and support were differentiated by department throughout, so a team further ahead in its AI use and a team just starting out each had a route that matched where they actually stood.

A Framework Built to Be Used

DTTT's starting position was that AI only becomes an organisational capability once it is applied consistently rather than left to individual initiative and the strategy was built around that belief. Every recommendation is judged against a simple, practical test: does it give staff better access to information, does it free up time lost to manual tasks and does it raise the consistency of quality across every team and every market? Foundations are established before new applications are added, internal processes are proven before anything external-facing is attempted and every department is supported from where it actually stands rather than an assumed common starting point. The result is a strategy built to earn trust through consistent, visible progress.

Central to that thinking is a governance model that scales oversight to the level of risk involved, rather than applying the same scrutiny to every piece of work regardless of its stakes. An AI Committee holds accountability across legal, data, compliance, capability and coordination, but its role is to set the standards instead of approving every decision. Internal work needs nothing more than good judgement, but work that is prepared for external communications requires a review from a team lead or even the Committee itself if there are strategic decisions about sensitive information. Structured this way, governance moves at the same pace as the organisation, which is what allows staff to trust it rather than work around it.

An AI strategy only holds if it lives inside an organisation's everyday network of people, not just its policy documents, which is why DTTT built a Committee, Champions and Ambassadors structure into the strategy. That structure is what allows AI to shift from simply automating isolated tasks, into augmenting how the organisation works, bringing new capability, better efficiency and clearer standards into everyday practice over time.

AI Champions carry that shift at ground level, spending a modest one to two hours a week building real expertise in a tool or workflow and passing what they learn to colleagues doing similar work, so useful approaches get shared across the organisation. AI Ambassadors then carry that same experience upward, giving leadership a current view of where AI is creating value and where it still needs support, so that decisions about capability are made systematically.

DTTT also brought five of its own frameworks into the strategy. The thinking behind this is that a strategy should leave an organisation with more than recommendations on a page, it should leave it with tools it can actually put to use. The transparency scale is a good example, giving A.T.A. a practical way to be open about where AI has shaped its work rather than just a principle to aspire to. The same thinking carries through the other frameworks, covering staff wellbeing, skills development and how individual pieces of work get reviewed before they go further, each one built to be picked up and adapted as A.T.A.'s own practice matures.

From Strategy to Daily Practice

The roadmap turns that structure into action across three phases. The first builds the foundations by publishing governance standards, launching the AI Champions network, ensuring the organisation's information is properly organised and delivering other quick wins early to build momentum. The second activates priority use cases and starts documenting how teams actually work. The third reaches into more strategic territory, including how Aruba is discovered and presented through AI search and how A.T.A.'s myAruba Assistant chatbot develops further. Internal work comes first throughout, a deliberate sequencing choice held even with a live external product already earning trust in the market.

Most of what success looks like here still lies ahead. Within a year, A.T.A. expects every department to have at least one documented AI-assisted workflow in regular use, confidence scores to have improved across the organisation and several cross-departmental use cases producing results teams can point to.

‍

More opinions
All opinions
AI

Enhancing AI's Role as a Channel for Experience Discovery

13 Nov 2026
AI

Earning Trust When "Real" is in Question

06 Oct 2026
AI

Why AI Security and Oversight Are Becoming Essential as AI Adoption Scales

23 Sep 2026

Industry research, open frameworks and field-tested practice on how destinations lead through AI, sustainability and brand. Since 2012.